(1995) originally developed a measurement tool for compassion fatigue which was called the Compassion Fatigue Stress Test. The tool was revised and renamed the Professional Quality of Life Scale by Stamm in 2005. Multiple revisions have taken place, but the final and most current scale, the ProQOL 5, is a thirty item questionnaire. Participants indicate, using a five-point Likert scale of (1 never, 2 rarely, 3 sometimes, 4 often, and 5 very often), how frequently they have experienced certain feelings in the last thirty days. The ProQOL 5 responses are divided into three separate 10-item subscales: compassion satisfaction, burnout, and secondary traumatic stress. Each subscale measures different constructs. The three subscales are scored separately. The burnout and secondary traumatic stress subscales represent the two components of
compassion fatigue and are interpreted as a combination to determine risk of compassion fatigue. The ProQOL 5 questions vary from “I am happy” to “I feel depressed because of the traumatic experiences of the people I help” (Stamm, 2010, p. 26).
The ProQOL 5 has been extensively tested and has been found to be reliable and valid (Hooper et al., 2010). According to the tool’s author, Stamm (2010), the construct validity is good with “more than 200 published papers and more than 100,000 articles on the internet,” (p.
30
13). Alpha reliability of the three scales is as follows: Compassion Satisfaction: alpha = 0.88; Burnout: alpha = 0.75; Secondary Traumatic Stress: alpha = 0.81. Between Compassion Fatigue and the Secondary Traumatic Stress scale there is a 2% shared variance (r=0.23; co-σ = 5%; n=1187). A 5% shared variance (r = -0.14; co-σ = 2%; n=1187) exists between Compassion Fatigue and the Burnout scale. The shared variance between the Burnout and Secondary Traumatic Stress scales is 34% (r = 0.58; co-σ = 34%; n = 1187), which is due to the scales measuring different constructs while reflecting distress inherent in both conditions (Stamm, 2010). Permission is given for the ProQOL 5 to be freely copied as long as the author is credited, no changes are made, and it is not sold is indicated in the footer of the instrument (Appendix E; Stamm, 2010).
Demographic Profile. The second tool used for data collection was a demographic
profile (Appendix D). The profile included categories as analyzed by Stamm (2010) utilizing a data bank of 1,289 cases from multiple studies. The profile included questions related to gender (male/female); age group (18-35, 36 and up); race (white, non-white); income group (up to $45,000 USD, $46,000-$75,000 USD, more than $75,000); years worked at current healthcare facility (<5 years, 5-15 years, >15 years); and years worked in critical care (<5 years, 5-15 years, >15 years), (Stamm, 2010). The profile also asked about length of shifts worked (eight hours, twelve hours, combination); type of shift worked (day, evening, night, combination); and number of hours worked each week (<24 hours, 24-36 hours, >36 hours).
Educational Module
All nurses working in the two critical care units were contacted by email and asked to participate in an online educational module (Appendix F) on the topic of compassion fatigue. A link to the online module was provided in the email. Confidentiality was protected by not asking
31
for any identifiers in order to complete the module. Instructions for completion as well as an overview of what to expect were provided in detail in the email. The module began with a pre- test to test compassion fatigue knowledge prior to the content. The content on compassion fatigue followed. A post-test to assess compassion fatigue knowledge after the content delivery was included. The last step was a survey for participants’ feedback on content and delivery, what they intended to implement, and perceived benefits of completing the module. Estimated time to complete the module was also provided. Results from the pre-tests, post-tests, and surveys were collected and analyzed.
To promote participation, reminder emails were sent asking participants to complete the module. Posters were hung in the units’ break rooms which encouraged nurses to participate. Nurse managers and clinical care supervisors from the units were given information related to the potential benefits of enhanced awareness of compassion fatigue on the units and were asked to announce the module to the nurses.
Data Analysis
Compassion Fatigue Survey
The three separate subscales of the ProQOL 5 (compassion satisfaction, burnout, secondary traumatic stress) were scored by the researcher using the ProQOL 5 scoring and interpretation tools (Appendix I, Appendix J). The responses of 1, 2, 3, 4, or 5 in the designated number slots for each subscale were summed, with the sum indicating the participant’s score for that particular subscale. Information by Stamm (2010) provided in Appendix J indicates what the score means and the level of compassion satisfaction, burnout, or secondary traumatic stress. Levels are reported as low, average, or high for each of the three subscales.
32
The average score for the compassion satisfaction subscale is 50 (standard deviation: 10, alpha scale reliability: 0.88). The average score for the burnout subscale is 50 (standard
deviation: 10, alpha scale reliability: 0.75). The average score for the secondary traumatic stress subscale is 50 (standard deviation: 10, alpha scale reliability: 0.81). Typically for each subscale participants complete, 25% score higher than 57, and 25% score lower than 43 (Stamm, 2010). Scores were coded and entered into SAS. Mean scores were calculated for each subscale within each demographic.
The statistical analysis of the data was performed with assistance from North Dakota State University Statistical Consulting Department. Descriptive statistics were used to analyze demographic information provided by participants and inferential statistics were used to explain the relationships between scores of subscale per demographic variables. T-tests were performed to find differences between scores of subscales by demographic variables with two levels. A post-hoc test using Fisher’s least significant difference (LSD) method was used to detect pairwise differences among the variables with three or more levels, after an initial one-way
analysis of variance (ANOVA).
Educational Module
Results from the pre-tests and post-tests were analyzed to determine how many
participants scored higher on the post-test than the pre-test. The participants’ narrative responses from the survey were analyzed for themes. The results from the pre-test/post-test analysis will be covered in the following chapter. The prominent themes from the survey responses will also be described in Chapter Four.
33